The oil addition would seem to explain the TBN. The engine uses ~1/2 quart per 1000 miles, but there is no oil added at some of the sampling intervals. Makup oil was added at 8000 miles, but not at 7,000. It would appear that addition of ~1/2 quart of fresh oil with 15 mg KOH equivalents per gram every 1000 miles or so just balances the amount of acid that forms over that interval. So once the TBN reached ~6 mg/gm, it appears to stabilize as the addition of make up oil raises the TBN at the beginning of the next interval. The rate of acid formation in the oil (which is really an oxidation reaction) slows down over time as the more easily oxidizable components in the oil react first. Of course, once the oxidation inhibitors in the additive package become fully depleted, the TBN is going to drop very fast. They you will KNOW the oil is finished.
BTW 3MP, this is a great experiment. I’d like to send you 20 bucks, but I can’t find you on the Cave Drawings link you supplied.